Development of multicellular organisms depends, at least in part, on mechanisms which specify, direct or maintain positional information to pattern cells, tissues, or organs. Various secreted signaling molecules, such as members of the transforming growth factor-beta (TGFβ), Wnt, fibroblast growth factors and hedgehog families have been associated with patterning activity of different cells and structures in Drosophila as well as in vertebrates. Perrimon, Cell 80:517-520 (1995).
Hedgehog (Hh), first identified as a segment-polarity gene by a genetic screen in Drosophila melanogaster (Nusslein-Volhard et al., Roux. Arch. Dev. Biol. 193: 267-282 (1984)), plays a wide variety of developmental functions (Perrimon, supra). Although only one Drosophila Hh gene has been identified, three mammalian Hh homologues have been isolated: Sonic Hh (SHh), Desert Hh (DHh) and Indian Hh (IHh) (Echelard et al., Cell 75: 1417-30 (1993); Riddle et al., Cell 75: 1401-16 (1993)). SHh is expressed at high level in the notochord and floor plate of developing vertebrate embryos. In vitro explant assays as well as ectopic expression of SHh in transgenic animals show that SHh plays a key role in neuronal tube patterning (Echelard et al., supra.; Ericson et al., Cell 81: 747-56 (1995); Marti et al., Nature 375: 322-5 (1995); Krauss et al., Cell 75, 1432-44 (1993); Riddle et al., Cell 75: 1401-16 (1993); Roelink et al., Cell 81:445-55 (1995); Hynes et al., Neuron 19: 15-26 (1997)). Hh also plays a role in the development of limbs (Krauss et al., Cell 75: 1431-44 (1993); Laufer et al., Cell 79, 993-1003 (1994)), somites (Fan and Tessier-Lavigne, Cell 79, 1175-86 (1994); Johnson et al., Cell 79: 1165-73 (1994)), lungs (Bellusci et al., Develop. 124: 53-63 (1997) and skin (Oro et al., Science 276: 817-21 (1997)). Likewise, IHh and DHh are involved in bone, gut and germinal cell development (Apelqvist et al., Curr. Biol. 7: 801-4 (1997); Bellusci et al., Dev. Suppl. 124: 53-63 (1997); Bitgood et al., Curr. Biol. 6: 298-304 (1996); Roberts et al., Development 121: 3163-74 (1995)). SHh knockout mice further strengthened the notion that SHh is critical to many aspect of vertebrate development (Chiang et al., Nature 383: 407-13 (1996)). These mice show defects in midline structures such as the notochord and the floor plate, absence of ventral cell types in neural tube, absence of distal limb structures, cyclopia, and absence of the spinal column and most of the ribs.
At the cell surface, the Hh signal is thought to be relayed by the 12 transmembrane domain protein Patched (Ptch) (Hooper and Scott, Cell 59: 751-65 (1989); Nakano et al., Nature 341: 508-13 (1989)) and the G-protein-coupled-like receptor Smoothened (Smo) (Alcedo et al., Cell 86: 221-232 (1996); van den Heuvel and Ingham, Nature 382: 547-551 (1996)). Both genetic and biochemical evidence support a receptor model where Ptch and Smo are part of a multicomponent receptor complex (Chen and Struhl, Cell 87: 553-63 (1996); Marigo et al., Nature 384: 176-9 (1996); Stone et al., Nature 384: 129-34 (1996)). Upon binding of Hh to Ptch, the normal inhibitory effect of Ptch on Smo is relieved, allowing Smo to transduce the Hh signal across the plasma membrane. Loss of function mutations in the Ptch gene have been identified in patients with the basal cell nevus syndrome (BCNS), a hereditary disease characterized by multiple basal cell carcinomas (BCCs). Disfunctional Ptch gene mutations have also been associated with a large percentage of sporadic basal cell carcinoma tumors (Chidambaram et al., Cancer Research 56: 4599-601 (1996); Gailani et al., Nature Genet. 14: 78-81 (1996); Hahn et al., Cell 85: 841-51 (1996); Johnson et al., Science 272: 1668-71 (1996); Unden et al., Cancer Res. 56: 4562-5; Wicking et al., Am. J. Hum. Genet. 60: 21-6 (1997)). Loss of Ptch function is thought to cause an uncontrolled Smo signaling in basal cell carcinoma. Similarly, activating Smo mutations have been identified in sporatic BCC tumors (Xie et al., Nature 391: 90-2 (1998)), emphasizing the role of Smo as the signaling subunit in the receptor complex for SHh. However, the exact mechanism by which Ptch controls Smo activity still has yet to be clarified.
Importantly, the signaling mechanisms by which the Hh signal is transmitted from its receptor to downstream targets also remain to be elucidated. Genetic epistatic analysis in Drosophila has identified several segment-polarity genes which appear to function as components of the Hh signal transduction pathway (Ingham, Curr. Opin. Genet. Dev. 5: 492-8 (1995); Perrimon, supra). These include a kinesin-like molecule Costal-2 (Cos-2) (Robbins et al., Cell 90: 225-34 (1997); Sisson et al., Cell 90: 235-45 (1997)), a protein designated fused (Preat et al., Genetics 135: 1047-62 (1990); Therond et al., Proc. Natl Acad Sci. USA 93: 4224-8 (1996)), and a zinc finger protein Ci. (Alexandre et al., Genes Dev. 10: 2003-13 (1996); Dominguez et al., Science 272: 1621-5 (1996); Orenic et al., Genes Dev. 4: 1053-67 (1990)). Additional elements implicated in Hh signaling include the transcription factor CBP [Akimaru et al., Nature 386: 735-738 (1997)], the negative regulator slimb [Jiang and Struhl, Nature 391: 493-496 (1998)] and the SHh response element COUP-TFII [Krishnan et al., Science 278: 1947-1950 (1997)]. In addition, a molecule designated Suppressor of fused (Pham et al., Genetics 140: 587-98 (1995); Preat, Genetics 132: 725-36 (1992)), found in Drosophila, is believed to be a component of the Hh signal transduction pathway.
Functional roles and interactions of these Hh pathway molecules have been suggested based in part on genetic and structural analyses. Mutants in Cos-2 are embryonicly lethal and display a phenotype similar to Hh over expression, including duplications of the central component of each segment and expansion domain of Hh responsive genes. In contrast, mutant embryos for fused and Ci show a phenotype similar to Hh loss of function, including deletion of the posterior part of each segment and replacement of a mirror-like image duplication of the anterior part or each segment and replacement of a mirror-like duplication of the anterior part (Busson et al., Roux. Arch. Dev. Biol. 197: 221-230 (1988)). Molecular characterizations of Ci suggested that it is a transcription factor which directly activates Hh responsive genes such as Wingless and Dpp (Alexandre et al. (1996), supra, Dominguez et al. (1996), supra). Likewise, molecular analysis of fused reveals that it is structurally related to serine threonine kinases and that an intact N-terminal kinase domain and a C-terminal regulatory region are required for its proper function (Preat et al., Nature 347: 87-9 (1990); Robbins et al., (1997), supra; Therond et al., Proc. Natl. Acad. Sci. USA 93: 4224-8 (1996)). Consistent with the putative opposing functions of Cos-2 and fused, fused mutations are suppressed by Cos-2 mutants and also by Suppressor of fused mutants (Preat et al., Genetics 135: 1047-62 (1993)). Whereas fused null mutations and N-terminal kinase domain mutations can be fully suppressed by Suppressor of fused mutations, C-terminus mutations of fused display a strong Cos-2 phenotype in a Suppressor of fused background. This suggests that the fused kinase domain can act as a constitutive activator of SHh signaling when Suppressor of Fused is not present. Recent studies have shown that the 92 kDa Drosophila fused, Cos-2 and Ci are present in a microtubule associated multiprotein complex and that Hh signaling leads to dissociation of this complex from microtubules (Robbins et al., Cell 90: 225-34 (1997); Sisson et al., Cell 90: 235-45 (1997)). Both fused and Cos-2 become phosphorylated in response to Hh treatment (Robbins et al., supra; Therond et al., Genetics 142: 1181-98 (1996)), but the kinase(s) responsible for this activity(ies) remain to be characterized.
To date, the only known vertebrate homologues for these components are members of the Gli protein family (e.g., Gli-1, Gli-2 and Gli-3). These are zinc finger putative transcription factors that are structurally related to Ci. Among these, Gli-1 was shown to be a candidate mediator of the SHh signal [Hynes et al., Neuron 15: 35-44 (1995), Lee et al., Development 124: 2537-52 (1997); Alexandre et al., Genes Dev. 10: 2003-13 (1996)] suggesting that the mechanism of gene activation in response to Hh can be conserved between fly and vertebrates. To determine whether other signaling components in the Hh cascade are evolutionarily conserved and to examine the function of fused in the Hh signaling cascade on the biochemical level, the human fused cDNA was isolated and characterized (see USSN provisional application 06/076,072, filed Feb. 26, 1998, which is incorporated herein in its entirety). In the mouse, fused is expressed in SHh responsive tissues. Biochemical studies demonstrate that fused is a functional kinase. Functional studies provide evidence that fused is an activator of Gli and that a dominant negative form of fused is capable of blocking SHh signaling in Xenopus embryos. Together these data demonstrated that both Cos-2 and fused are directly involved in Hh signaling.
Recently, in Drosophila, a suppressor of the fused protein has been identified and shown to be a novel PEST-containing protein (Monnier et al., Curr. Biol. 8:583-586 (1998), Pham et al., Genetics 140:587-598 (1995), Preat et al., Genetics 135:1047-1062 (1993) and Preat, Genetics 132:725-736 (1992)). PEST domains are short sequences enriched in proline, glutamic acid (or aspartic acid), serine and threonine (single letter codes P, E, S, and T respectively), combined with a low hydrophobicity index. They are found in many proteins with short (<2 hour) cellular half-lives (40). Applicants have herein identified and described a DNA encoding a polypeptide having homology to that suppressor polypeptide and designated herein as human Suppressor of fused (“hSu(fu)”), and alternatively as hSu(fu). Somatically acquired mutations of the patched gene have been identified in sporadic cancers, including basal cell carcinomas, primary breast carcinomas, medulloblastomas and meningiomas. It is currently believed that patched acts as a tumor suppressor, and that these mutations cause a loss of function in the patched gene product. The hedgehog/patched signaling pathway may therefore be a factor in tumorigenesis. Detecting genetic alterations that lead to increased cell growth and tumorigenesis is of great interest for clinical medicine. Identifying the specific changes that lead to altered cell growth may open the door to improved diagnosis and possible treatment for associated tumors.